Nabtesco

Phase A — Understand the business

Lens 1 · Company Overview

Nabtesco makes the RV cycloidal precision reduction gear that sits in the base, shoulder and elbow joints of medium-and-large industrial robots and converts a fast, weak servo motor into a slow, immensely strong, near-zero-backlash rotary output. In the same way Harmonic Drive is the strain-wave gear, Nabtesco is the RV reducer — and it is the larger, more diversified company of the two. Management states an ~60% global share in medium-to-large robot joints, with third-party estimates that Nabtesco alone supplies >35% of the RV units used in articulated robots worldwide and, on the broadest framing, "reducers for ~60% of global industrial-robot output". This is a picks-and-shovels toll-booth on factory automation — and, the market now hopes, on humanoids.

But Nabtesco is not a pure-play reducer company, and that is the whole investment distinction versus Harmonic Drive. After the FY2025 divestiture of its low-margin Hydraulic Equipment business (construction-machinery travel motors — ironically the origin of the RV gear, which began as an excavator travel-drive), the company reports three continuing segments:

• Component Solutions — ¥79.3B revenue (~25.8% of continuing sales). The crown jewel: precision RV reducers for industrial robots + gears for construction/agricultural machinery. This is the segment the humanoid narrative attaches to.
• Transport Solutions — ¥100.5B revenue (~32.6%), OP ~¥13.6B (~13.5% margin). Railway brake systems and door devices, marine control, and commercial-aircraft flight-control actuators (Boeing 777X primary flight control, 787 power-distribution, near-100% share of Japan-built aircraft flight-control actuation). Long-cycle, high-barrier, aftermarket-heavy — the ballast.
• Accessibility Solutions — ¥110.7B revenue (~36.0%), OP ~¥9.1B (~8.2% margin). Automatic doors (the Nabco / Gilgen platform-door and building-door franchise), platform safety gates, welfare/care equipment. The largest segment by sales, steadiest demand, thinnest margin.

Total FY2025 continuing net sales ¥307.9B (+9.8% YoY), with the balance in "Manufacturing Solutions / Others". So robots are the story, but roughly three-quarters of revenue is rail brakes, aircraft actuators and automatic doors — a very different, more defensive company than the market's "humanoid reducer play" framing implies. That gap is the central tension of this dossier (bull and bear both live in it).

Contract structure. Component sales are book-and-ship purchase orders — order-driven, cyclical, levered to robot-OEM and construction capex, no annuity (Lens 5's ¥98.3B Q1 order figure is the tell, not backlog contracts). But Transport (aerospace/rail) carries decade-long platform lives + a sticky, high-margin spares/MRO annuity, and Accessibility (doors) carries installed-base maintenance. Blended, Nabtesco has far more recurring, non-cyclical revenue than Harmonic Drive — the diversification the pure-play lacks.

Commercial-layer anchor: the KB robotics map (kb/robotics/wiki/ pointers exist but the wiki files are not yet populated — positioning.md/bottlenecks.md/supply-chain.md return not-found) names the "Actuators / motors → joint torque" node and "high-torque compact actuators" as a supply chokepoint held by "Harmonic Drive + emerging Chinese." Nabtesco is the co-node on that chokepoint for the heavy (RV) side; this dossier is the company-level fill.

Lens 2 · Supply Chain

Upstream → Nabtesco → end customer, named:

• Upstream inputs: specialty bearing steels and case-hardening alloys; precision grinding/hobbing and heat-treatment; the RV gear's binding step is the eccentric cam + cycloidal disc + needle-roller assembly machined to single-digit-micron tolerance — durability under shock load is the guarded competence (inherited from the excavator-travel-motor lineage). Servo motors/encoders sourced for the mechatronic sub-assemblies. Japan-centric core machining; a Changzhou, China production base (Changzhou Nabtesco Kusaka Precision Machinery) does local gear-wheel manufacture — RMB 583M output in 2024.
• Nabtesco (the node): centralised precision machining in Japan (guards the process), localised assembly/finishing near customers (Japan, China, and the acquired European door/rail footprint). Reducer capacity doubling by 2026 is the pre-emptive humanoid build.
• Downstream customers (named):
• Industrial-robot OEMs — Fanuc (6954.T), Yaskawa (6506.T), ABB, KUKA, Kawasaki — the big-4-plus that put RV gears in every 6-axis arm base/shoulder.
• Construction/ag machinery — travel and swing drives (the legacy application).
• Aircraft OEMs — Boeing (777X, 787, 737/747-8/757/767 flight-control actuation), plus Japanese aerostructures primes; rail operators/rolling-stock builders for brakes and doors; building/infra for automatic doors and platform gates.
• Humanoid developers — the claimed future pipeline. CRITICAL CAVEAT (Lens 13): the flagship humanoids overwhelmingly use strain-wave (harmonic) reducers, not RV — Tesla Optimus uses harmonic reducers; the marginal humanoid win is going to Harmonic Drive and to Zhejiang Shuanghuan (which supplies both Optimus and Unitree G1). Nabtesco's RV is designed for heavy joints; its humanoid entry is the new compact RVmini / Monocrank line launched Dec 2025, not the core RV. This is the single most important supply-chain fact in the whole file.

Chokepoints & single-source dependencies:

The asymmetry: Nabtesco is the chokepoint for heavy industrial-robot joints and for Japan aerospace actuation — genuinely defensible. Its exposure is (1) a good-enough Chinese RV substitute qualifying in at the low end, and (2) being on the wrong side of the reducer-type split for the humanoid growth vector everyone is paying for.

Lens 3 · Competitive Advantages (moats)

A genuine wide moat on the base business — Morningstar explicitly rates Nabtesco "wide-moat" — layered over two very different franchises.

1. RV process IP + design-in switching costs (Component). Decades of shock-load-durable cycloidal metallurgy; ~60% share concentrated in the premium/heavy joint tier where field failure is catastrophic. A robot arm is engineered around a specific reducer's stiffness/backlash/torque curve; re-qualifying a second source forces re-tuning the motion/control policy — a real, if not impregnable, switch cost.
2. Aerospace certification + platform-life lock (Transport). Flight-control actuators are certified per airframe and locked for the platform's 20–40-year life with a captive, high-margin spares/MRO annuity. This is a TransDigm-style aftermarket moat that Harmonic Drive completely lacks — and it is invisible in the "humanoid reducer" framing.
3. Installed-base maintenance (Accessibility). Automatic-door and platform-gate installed base pulls recurring service revenue; brand/reliability incumbency in infrastructure.
4. Scale + diversification. ~¥308B revenue across three weakly-correlated end-markets (factory capex / air travel + rail / building infrastructure) — a materially more resilient earnings base than the single-cycle pure-plays.

Bargaining power — the honest read. Strong over Western/Japanese robot OEMs and over Japanese aircraft programs (often the only qualified source). Structurally weaker in China, where domestic RV/harmonic localisation (MIIT-backed Shuanghuan; Leaderdrive at ~15% global / ~26% domestic share by 2023 ) is displacing imported reducers on price. And weakest of all in the humanoid vector, where Nabtesco's dominant product isn't the preferred geometry. The durable moat is real; the growth moat the multiple is pricing is the shakiest leg.

Lens 4 · Segments

By segment:

Consolidated FY2025: net sales ¥307.9B (+9.8%), operating income +60.3% YoY (a recovery off a depressed FY2024 base — implied OP ≈ ¥20.7B, ~6.7% consolidated margin ), net income ¥15,695M (+55.1%), EPS ¥131.56, dividend ¥80.

The story the segments tell. The narrative is 100% Component/robots, but Transport (aerospace + rail) is the profit engine at ~13.5% margin, and Accessibility is the revenue anchor. Component's margin was compressed in the downcycle and the entire operating-leverage recovery (FY2025 OP +60%, FY2026 guide +33.6%) is the reducer business re-absorbing fixed cost as robot capex recovers and price pass-through sticks (Lens 5). So the segment that carries the valuation (Component) is the one whose margin is most cyclical and whose humanoid upside is most contestable — while the segments that carry the earnings (Transport/Accessibility) get almost no credit in the multiple. That mismatch is the crux. (Segment mix + Component margin recovery are the moving parts — re-run on refresh.)

---

Phase B — Measure performance

Lens 5 · Earnings Result

Two prints matter: FY2025 (year to Mar-2026) and the very strong Q1 FY2026 (Jun-2026).

FY2025 (reported ~mid-May 2026):

• Net sales ¥307.9B, +9.8% YoY — a hair below the company's own top-line forecast but profit surged above forecast.
• Operating income +60.3% YoY (≈¥20.7B, ~6.7% consolidated margin ) — a sharp recovery off the FY2024 trough, driven by robot-capex recovery, price pass-through on reducers, and the mix-cleanup from exiting hydraulics.
• Net income ¥15,695M, +55.1%; EPS ¥131.56; ROE strengthened.
• Dividend ¥80/sh, plus a ¥10B / up-to-4.0M-share (3.3% of shares) buyback — the capital-return signal a TSE-reform-era Japanese industrial is expected to send.

Q1 FY2026 (quarter to Jun-2026, reported ~early Aug 2026) — a clean beat-and-raise:

• Net sales ¥83.0B, +16.9% YoY; operating profit ¥8.2B, +68% YoY; operating margin 9.9% (+3.0pts).
• Orders received ¥98,257M, +28.4% YoY, up across all segments — the leading indicator, and it's accelerating.
• Net income ¥5,483M, +67.0%; EPS ¥46.78 (vs ¥27.33).
• Sales-growth drivers: Precision Reduction Gears +¥4.8B, Automatic Doors +¥4.0B, Marine +¥2.2B — i.e. broad, not a single-line spike.
• Guidance raised: FY2026 net sales ¥327.0B, operating income ¥27.7B (+33.6%), net income ¥18.6B, EPS ¥158.72, dividend ¥82.
• Balance sheet: equity ratio 61.7%, cash ¥80.6B — conservative, net-cash-like.

Read: unlike Harmonic Drive (which guided a huge recovery but printed a falling net income on unabsorbed capex), Nabtesco is already delivering the operating leverage in the actuals — OP +60% then +68%, margin visibly climbing toward the 11% mid-term target, orders +28%. The recovery is shown, not just guided. That is the single strongest fact in the bull case and the main reason the multiple (Lens 7) is a fraction of Harmonic's. Market reaction: the stock is a top humanoid-thematic name, +46% YTD and +141% 1-year TSR into mid-2026 — it has re-rated, but off rising real earnings, not purely off narrative.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the shelf (web-only; Roic.ai lists 6268.T transcripts but they weren't ingested). From briefing coverage the tonal arc across the last ~4 quarters:

• FY2024 / early FY2025 (trough): defensive — weak robot + construction capex, China reducer softness, compressed Component margin; the drag that justified exiting hydraulics.
• Mid-FY2025: the pivot — portfolio reform executed (hydraulics divested), framing shifts to "focus on higher-margin precision motion + price pass-through + utilization recovery," and the humanoid-capacity-doubling message enters.
• FY2025 results → Q1 FY2026 (May–Aug 2026): confident and raising — OP +60% then +68%, orders +28%, guidance lifted, buyback announced, capacity doubling on track, new RVmini/Monocrank compact gears (Dec-2025) explicitly aimed at collaborative/humanoid robots.

What they added: "price pass-through," "utilization recovery to 11% margin," "portfolio reform," "compact precision reduction gears for new applications (cobots/humanoids)." What they stopped saying: the hydraulics-drag / cyclical-apology framing. Sentiment shift: decisively positive — but more operationally grounded and less promotional than Harmonic Drive's (Nabtesco is pointing at delivered margin recovery and orders; Harmonic is pointing at a 2030 option). That relative sobriety is a mild governance positive.

Lens 7 · Comps

Peer set: the two Japanese reducer leaders (Nabtesco RV, Harmonic strain-wave), the automation platform (Fanuc), the Chinese challengers (Estun, Leaderdrive/Leader-Harmonious, Shuanghuan), and the motion-adjacent (HIWIN).

Read — this is the entire investment case in one table. Nabtesco and Harmonic Drive are the two Japanese reducer champions riding the same humanoid thematic (KOID weights: Harmonic 2.31%, Nabtesco 2.21%, Leaderdrive 1.94% ). Yet Nabtesco trades at ~2.1x sales / ~39x earnings vs Harmonic's ~10.4x sales / ~186x — a ~5x sales-multiple discount — despite being larger, more diversified, more profitable today (rising OP vs falling), and Morningstar-rated wide-moat. On the surface Nabtesco is the vastly more sane way to own the reducer duopoly. The catch the discount is correctly pricing (Lens 13): Nabtesco's RV geometry is the wrong gear for the humanoid vector that justifies Harmonic's premium — so Harmonic gets the option value and Nabtesco gets the cyclical-industrial multiple plus a modest humanoid kicker. Against Fanuc (~39x fwd, but ~27% EBITDA margin and ~19% net margin), Nabtesco's ~39x on a ~7–10% consolidated margin is not cheap on quality-adjusted earnings — it is cheap only relative to Harmonic, and fair-to-full on its own fundamentals. Simply Wall St pegs fair value ~¥4,135 vs the ~¥5,655 price → ~14–37% overvalued on DCF across their notes.

Lens 8 · Stock-Price Catalysts (moves >5%, ~last 5 years — pattern)

Web-sourced; the pattern matters more than any single date. 52-week range ¥2,536–¥6,113, all-time high ¥6,113 on 14 May 2026, ~¥5,655 on 3 Jul 2026 — a ~2.4x move off the low.

• 2024 downcycle de-rate: weak robot/construction capex + China reducer share loss + hydraulics drag → the stock based near ¥2,536.
• Portfolio-reform + humanoid-theme re-rate (2025→2026): hydraulics divestiture, "physical AI / humanoid reducer" thematic inclusion (KOID and peers), capacity-doubling announcement → the bulk of the +141% 1-year move.
• FY2025 results (OP +60%, buyback) + Q1 FY2026 beat-and-raise (OP +68%, orders +28%, guide up): earnings actually delivering the recovery → sustained bid to the ¥6,113 ATH.
• RVmini/Monocrank launch (Dec-2025) + iREX 2025 / automatica 2025 robot-show presence: product proof-points for the cobot/humanoid entry.
• Thematic beta: moves with Tesla Optimus / physical-AI headlines and Nikkei robot-sector sentiment; broker consensus Buy, avg target ¥5,154 (high ¥6,300 / low ¥3,200, 11 analysts) — i.e. the stock is slightly above consensus fair value.

What the market actually reacts to: (1) the robot-capex cycle + orders — the real current earnings driver, and it's inflecting up; (2) humanoid-thematic flows (KOID/basket beta) disproportionate to Nabtesco's actual humanoid revenue; (3) portfolio-reform/capital-return proof-points (divestiture, buyback); (4) broker targets on a thinner Japanese float. It reacts less to the China-RV-erosion and the reducer-geometry mismatch — the two risks the tape is under-weighting.

---

Phase C — Judge people & books

Lens 9 · Management

• President & CEO: Kazumasa Kimura (since 2022). A new management structure + officer changes were announced 28 Nov 2025 — governance refresh consistent with the portfolio-reform push; the exact incoming line-up is not cleanly sourced here (flag for a Japanese-filing pull).
• Track record. The recent chapter is a genuine, well-executed portfolio-reform story: exited low-margin hydraulics, refocused on high-margin precision motion + aerospace/rail, and delivered the margin recovery (OP +60% then +68%, orders +28%, margin climbing toward the 11% target). That is real capital-discipline execution, not a slide-deck promise — the differentiator vs Harmonic's still-unproven leverage. The knock: they let Component margin and China share slip through 2024 before acting, and the multi-year franchise (aerospace/rail) is steady-state rather than a growth engine.
• Skin in the game. Institutional-owned (~47% institutions; top holders Eastspring ~5.3%, Fidelity ~5.1%, Nomura AM ~5.1%; public/retail ~42%), low cross-shareholdings relative to peers — a cleaner, more shareholder-aligned register than a typical old-line Japanese industrial. Individual insider ownership is (as usual for a large Japanese corporate) small and not the alignment mechanism — governance + institutional pressure is.
• Capital-allocation history. Conservative-but-improving and clearly responsive to TSE-reform pressure: net-cash-like balance sheet (equity ratio ~62%), ¥10B / 3.3% buyback, dividend ¥80→¥82, and counter-cyclical capacity investment (reducer capacity doubling ahead of demand). The mid-term plan targets ROIC >10% and operating-margin recovery to 11% in FY2026 via price pass-through, cost reduction and utilization recovery, plus a 50% CO2 cut by 2027. The read: prudent, reform-minded, returning cash — a better capital-allocation story than Harmonic's over-capitalised 2% ROE.
• Founder vs professional-manager archetype. Fully professional-manager (Nabtesco is a 2003 merger of Teijin Seiki + NABCO; no dominant founder-family). Implication: disciplined, governance-driven, TSE-reform-responsive — but likely slower and less aggressive on the China price war and on forcing into the strain-wave/humanoid geometry than a founder-led Chinese challenger.

Lens 10 · Forensic Red Flags

Web-only forensic read (no filings to tie out; every figure labeled):

• Earnings quality — no obvious red flag. OP and net income are rising on rising sales with margin expansion and a +28% order book — the opposite of the "sales up, profit down" pattern that flags trouble. Cash position ¥80.6B against equity ratio ~62%. Full operating-cash-flow-vs-net-income tie-out not sourced at line-item level (no filing on shelf) — flag as an open verification item, but nothing in summaries suggests accrual games.
• Segment OP granularity is the main data gap. Component Solutions operating profit/margin is not cleanly and independently sourced (the official briefing PDF won't parse); I reconstruct it as "recovering toward the 11% target." A reader underwriting the operating-leverage thesis should pull the JGAAP segment note directly to confirm Component margin trajectory — this is the number the whole re-rating rests on.
• Receivables/inventory vs revenue: not sourced at line-item granularity — flag.
• Balance-sheet quality: conservative (net-cash-like, equity ~62%); the divested hydraulics + acquired European door/rail businesses mean some goodwill/intangibles exist, but no evidence of a goodwill balloon in available data.
• SBC / non-GAAP flattering: not a factor — Japanese GAAP, no material stock-comp add-back culture.
• China subsidiary opacity: Changzhou operations are small (RMB 583M output 2024) and not individually broken out — low materiality, but the China demand erosion (not accounting) is the real risk (Lens 13).

Regulatory findings (required sub-section).

• SEC (EDGAR): No CIK — Nabtesco is not an SEC registrant, so no EDGAR LR/AAER search is possible. (Note: the historical SEC EDGAR hits under "Nabtesco Corporation and NABCO DOOR Ltd." relate to a 2012 Gilgen/door-business acquisition tender filing, not an enforcement action — do not attribute as a red flag.)
• Non-SEC (FTC/DOJ/antitrust/recall/JFTC): targeted web search "Nabtesco" (FTC OR DOJ OR fine OR penalty OR consent decree OR settlement OR antitrust OR recall) returned no material enforcement action against 6268.T as of 2026-07-07.
• Item 3 / legal proceedings: no 10-K on shelf (foreign filer); Japanese-GAAP securities-report legal-proceedings section not ingested — not sourced.
• Conclusion: No material regulatory or legal findings identified — verified via SEC EDGAR EFTS (no CIK, 0 findings) and targeted web search as of 2026-07-07. Japanese statutory-filing legal-proceedings disclosure remains unread and is an open verification item.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (EPS, next 3 fiscal years)

Built bottom-up from the FY2025 actual + Q1 FY2026 print + company guidance. Shares ~117.6M. Current price ¥5,655 (2026-07-03). No forecast.ts create in this unattended watchlist run — logged here as an estimate only.

Anchors: FY2025 actual EPS ¥131.56. FY2026 company guidance (raised) net income ¥18,600M → EPS ¥158.72.

Implied P/E at ¥5,655: Base FY2026 ~35.6x, FY2027 ~31x, FY2028 ~28x. On the bull path (FY2028 ¥250) the stock is ~23x three years out; on the bear path it's ~51x+ on falling numbers. Contrast with Harmonic Drive: ~186x guided vs Nabtesco's ~36x. Nabtesco's valuation is demanding but survivable — it prices continued cyclical recovery + modest humanoid optionality, not a decade of flawless humanoid volume. This is the mathematically defensible member of the reducer duopoly: you can lose the humanoid call and still own a ~13.5%-margin aerospace/rail franchise + a wide-moat RV near-monopoly at ~28–36x, whereas losing the humanoid call at Harmonic's ~186x is a 60–70% drawdown.

Brier-scoreable base call (not logged this run): "6268.T FY2026 (year ending 2027-03-31) net income ≥ ¥18,600M (meets raised guidance), p ≈ 0.62." Rationale: the Q1 beat-and-raise (OP +68%, orders +28%) gives real support, and the guide is already raised off a strong actual — higher confidence than Harmonic's un-shown leverage — but a robot-capex wobble or China acceleration remains the slip risk.

Lens 12 · Bull vs Bear

Bull case (narrative). Nabtesco is the sane, diversified toll-booth on physical automation — and the market is handing it to you at a fifth of the multiple it puts on its pure-play cousin. It owns ~60% of the medium/large robot-joint reducer market (Morningstar-rated wide-moat), and it's proving the operating leverage in real time: OP +60% then +68%, orders +28%, margin climbing to an 11% target, all off a portfolio already cleaned up by divesting low-margin hydraulics. Underneath the robot story sits a franchise Harmonic Drive can only dream of — Boeing 777X/787 flight-control actuation with a 20–40-year platform-life aftermarket annuity, plus rail brakes and an automatic-door installed base — three weakly-correlated end-markets that make the earnings base genuinely resilient. Management is returning cash (¥10B buyback, rising dividend) under TSE-reform pressure with a clean institutional register and ROIC >10% ambition. And there's a free option: the new RVmini/Monocrank compact gears aim Nabtesco's process moat at the cobot/humanoid market where, if physical AI scales, it captures upside it isn't being charged for at ~36x. You own a wide-moat, cash-returning, profitable near-monopoly with real cyclical recovery and a humanoid kicker — for a mid-30s P/E, not a 186x lottery ticket.

Bear case (2–3 permanent-impairment risks).

1. The reducer-geometry mismatch is fatal to the growth thesis. The humanoid re-rating everyone is paying for runs on strain-wave, not RV — Tesla Optimus uses harmonic reducers; Shuanghuan and Harmonic/Leaderdrive are winning the humanoid sockets. Nabtesco's crown-jewel RV is the heavy-industrial gear; its humanoid entry (RVmini, launched Dec-2025) is late and unproven. If humanoids are the growth story, Nabtesco is largely a spectator holding the wrong SKU — and a ~36x multiple that assumes it participates re-rates toward a cyclical-industrial ~20x.
2. China is commoditising the base business. MIIT-backed Shuanghuan and Leaderdrive (~15% global / ~26% China share by 2023, and rising) are localising RV and harmonic on price; imported-reducer share in Chinese robots is structurally falling. If that good-enough substitute walks up-market (cost is the gating constraint on cheap robots), Nabtesco's ~60% premium share erodes structurally, not cyclically — exactly the leak Harmonic already shows in its −28% China revenue.
3. The cycle is doing the heavy lifting, and it's a cycle. The +60%/+68% OP surge is an off-a-trough recovery on a book-and-ship, capex-levered order book. A 2027 robot/construction-capex rollover takes Component margin straight back down; aerospace/rail steady the ship but don't grow into the gap. At ~36x you're paying an above-mid-cycle multiple near a cyclical peak in the reducer recovery.

Pre-mortem (it's Jan 2028, thesis broke — what happened?). The robot-capex upcycle peaked in 2026–27 and rolled; Component margin round-tripped below 9%. The humanoid volume that did materialise went to strain-wave suppliers and Chinese RV, and Nabtesco's RVmini stayed a rounding error. A marquee builder headline ("in-house/Chinese-sourced actuator") de-rated the whole reducer basket. FY2026's raised guide was the peak. The stock re-rated from ~36x toward a cyclical-industrial ~18–20x on lower forward numbers — a ~40–50% drawdown — cushioned (unlike Harmonic) by the aerospace/rail/door earnings that never went away.

Are multiples too high? Relative to Harmonic Drive, no — Nabtesco is dramatically cheaper and better-quality. On its own fundamentals, mildly full: ~36x fwd on a ~6–10% consolidated-margin cyclical industrial, above Simply Wall St's ~¥4,135 DCF fair value. Defensible on the recovery + diversification; stretched if you also demand it wins humanoids.

Contrarian view (what the market refuses to see). Consensus treats Nabtesco and Harmonic Drive as two ways to play the same humanoid trade, and prices Nabtesco as the "cheap one." The thing the tape is missing is that they are not the same trade at all — Harmonic is the humanoid gear (strain-wave) and deserves an option multiple; Nabtesco is the industrial-robot + aerospace gear wearing a humanoid costume. The contrarian read cuts both ways: bulls over-credit Nabtesco's humanoid participation (it holds the wrong geometry), and bears under-credit its aerospace/rail annuity + wide-moat RV base (which is worth far more than a spectator's option). The right frame isn't "cheap humanoid play" — it's "wide-moat diversified industrial with a real cyclical recovery, a genuine aerospace aftermarket moat, and a lottery ticket on RVmini — priced as if the lottery ticket is a sure thing."

Lens 13 · Devil's Advocate (short-seller)

Skeptical short-seller dismantling the bull case.

• What structurally breaks the money machine? The premise that Nabtesco rides the humanoid wave. It doesn't — its dominant product (RV cycloidal) is not the humanoid reducer of choice (strain-wave). The single most-cited humanoid, Tesla Optimus, uses harmonic reducers; Shuanghuan supplies Optimus + Unitree; the humanoid TAM is being captured by strain-wave and Chinese players. Strip the humanoid premium and you have a mid-30s-P/E cyclical industrial — that's a de-rating, not a thesis.
• Where is revenue concentrated / what if concentration shifts? Component (robots) carries the valuation; a handful of robot OEMs (Fanuc, Yaskawa, ABB, KUKA) carry Component. If those OEMs (a) hit a capex downcycle or (b) dual-source Chinese RV to cut cost, Component margin collapses back to trough. Meanwhile the profit concentration in Transport (aerospace/rail) is steady but slow — it can't grow into a robot-cycle air-pocket.
• Why is the moat weaker than bulls think? In the premium/heavy tier it's strong. But moats are priced at the margin, and the margin is (1) China localisation eroding the base at ~40–60% of price, and (2) the humanoid geometry Nabtesco doesn't lead. A wide-moat rating on the 2024 franchise can be true while the 2027 growth is captured by others.
• Most dangerous competitor bulls underestimate? Zhejiang Shuanghuan — MIIT-backed, already in Optimus + Unitree, localising both RV and harmonic. It doesn't need to beat Nabtesco on premium reliability; it needs to be good-enough-and-cheaper for cost-obsessed volume robots, and it already is in China. Plus Harmonic Drive itself owning the humanoid socket Nabtesco is trying to enter late.
• Worst capital-allocation / governance marks? Actually few — the reform (hydraulics exit, buyback, ROIC target) is a positive. The fairest short knock is slowness: a professional-manager, consensus-driven Japanese house is structurally slow to price-fight China or force into strain-wave — it will defend the premium tier while the volume tier migrates.
• What must hold for ¥5,655? Continued robot-capex recovery, Component margin sustaining ~11%, buyback support, and enough humanoid/thematic flow to hold a ~36x multiple on a cyclical. Miss any one and consensus fair value (¥4,135–¥5,154) reasserts.
• Valuation if growth disappoints 20–30%? FY2026 EPS to ~¥120 (bear) at a de-rated ~20x cyclical multiple → ~¥2,400 — a ~55% downside to the trough range, but floored well above zero by the aerospace/rail/door earnings (the key contrast with Harmonic's uncushioned 186x).
• Single scenario that permanently impairs the business? Not one — the diversification means no single blow is fatal (that's the point). The realistic impairment-of-the-thesis (not the business) is: humanoids scale on strain-wave/China + a robot-capex rollover, and Nabtesco reverts to a well-run, wide-moat, ~18–20x cyclical industrial. That's a painful de-rate, not a bankruptcy — which is exactly why Nabtesco is a lower-conviction-but-lower-risk name than the pure-plays.

Lens 14 · Management Questions (ordered by information value)

1. Component Solutions operating margin has been the trough-to-recovery swing factor — what was Component's exact operating margin in FY2025 and Q1 FY2026, and what gets it durably to and above the 11% mid-term target versus mean-reverting with the robot cycle?
2. Tesla Optimus and most flagship humanoids specify strain-wave, not RV. What is Nabtesco's concrete plan and timeline to win humanoid-joint content — is RVmini/Monocrank a real strain-wave-competitive entry, and do you have named humanoid design wins or shipping volume today?
3. What is your current China RV-reducer unit share and how has it moved over the last three years against Shuanghuan and Leaderdrive — and where does it settle in three years?
4. How much of the FY2025/FY2026 operating-leverage surge is durable structural margin versus a cyclical off-the-trough recovery that reverses when robot capex normalises?
5. On price pass-through: how much reducer price increase has actually stuck with robot OEMs, and how do you defend price as Chinese RV closes the quality gap at 40–60% of your price?
6. What is Component's revenue concentration among the top robot-OEM customers, and what is your exposure if one dual-sources Chinese reducers to cut BOM cost?
7. Reducer capacity is doubling by 2026 — what utilization and demand assumptions underwrite that capex, and what is the margin drag if humanoid/robot volume undershoots?
8. Transport Solutions runs the highest segment margin — how large and how sticky is the aerospace/rail aftermarket/MRO annuity, and what is the 777X/787 content ramp over the next five years?
9. What is your through-cycle ROIC target by segment, and how do you justify holding the lower-margin Accessibility (door) business rather than divesting it as you did hydraulics?
10. Post-hydraulics divestiture, what is the next portfolio-reform move — are there other sub-scale or low-return businesses on the block?
11. How do you weigh buybacks/dividends versus M&A into strain-wave or humanoid-actuation capability given the net-cash balance sheet and TSE-reform pressure?
12. What is your assumption for the robot-capex cycle over FY2026–FY2028, and how does the order book (+28% in Q1) translate to shipments if macro softens?
13. Which end-market — factory automation, aerospace, rail, doors, or humanoid — do you expect to be the largest incremental profit contributor in three years, and why?
14. What technology shift (integrated actuators, direct-drive, in-house OEM reducers, a new gear geometry) is the biggest threat to the RV franchise, and how are you hedged?
15. What did the November 2025 management restructuring change about strategy execution, capital allocation, or the China/humanoid response speed?

---